Objective: Social media has been implicated as a leading driver of the youth vaping epidemic in the United States. Despite the recent proliferation of synthetic nicotine products in the marketplace, there is limited understanding of the promotion, health risks, and regulatory policy associated with these products. We aim to identify and characterize posts on Instagram and Twitter related to the promotion of synthetic nicotine products, self-reporting of adverse events following synthetic nicotine product use, and also identify discussion topics related to the regulation, health policy, and education about synthetic nicotine.
Methods: We conducted a hashtag and keyword search on Instagram and Twitter, respectively, to collect posts related to synthetic nicotine products. We then analyzed this data utilizing multilanguage BERT analysis, a pre-trained supervised topic modeling algorithm, to sort the dataset into clusters grouped based on textual similarity. After this, we manually annotated the most representative posts corresponding to each topic cluster using a codebook associated with characteristics of interest and categorized clusters by their coherence to themes of promotion, adverse events, and regulation.
Results: A total of 14,651 Instagram posts and 24,081 Twitter posts were collected from the keyword search. After an intermediary data cleaning phase to remove posts which could not be recognized by topic modeling, the multilanguage BERT topic model thematically clustered 49.4% (n=6034) of Instagram posts and 46.0% (n=3200) of tweets. After manual content analysis, we detected 52.9% (n=3193) of Instagram posts and 30.4% (n=972) of tweets that were in clusters thematically related to our study aims. The most representative theme on Instagram was synthetic nicotine electronic nicotine delivery systems (ENDS) promotion, with 91.6% (n=2924) of posts belong to that thematic cluster, followed by regulation and policy related posts (8.4%, n=268). In comparison, the most representative theme on Twitter related to self-reporting of adverse events (50.2%, n = 488), followed by promotion (39.8%, n=387), and regulation and policy (10.0%, n=97). Manual annotation of the most representative tweets within these clusters showed a higher level of coherence that Instagram clusters had towards its respective theme than Twitter clusters. A qualitative sub-analysis found that most of the synthetic nicotine ENDS promotion activity was more specifically related to the selling of synthetic nicotine ENDS products by different vendors.
Conclusion: Despite platform prohibitions against the marketing and sale of various tobacco products, there remains significant user-generated content related to synthetic nicotine products on Instagram and Twitter, with most posts related to the promotion and sales of synthetic nicotine ENDS products. Most of the synthetic nicotine ENDS content in our dataset on Instagram is closely related to the theme of ENDS product promotion, with little discussion of regulation and adverse events. On Twitter, synthetic nicotine ENDS content is more heterogeneous, with significant discussion of adverse events following synthetic nicotine ENDS use along with similar promotion of synthetic nicotine ENDS products. Further research is needed to better understand the acute health risks unique to synthetic nicotine products and whether or not these public health challenges are exacerbated due to unregulated and illegal promotion and sale of these products via social media. / 2025-03-08T00:00:00Z
Identifer | oai:union.ndltd.org:bu.edu/oai:open.bu.edu:2144/48374 |
Date | 08 March 2024 |
Creators | Shah, Neal |
Contributors | Gerstenfeld, Louis C., Mackey, Timothy K. |
Source Sets | Boston University |
Language | en_US |
Detected Language | English |
Type | Thesis/Dissertation |
Page generated in 0.0012 seconds